[Libreoffice-ux-advise] [Bug 135501] Change the default UI (see comment 67)

[bugzilla-daemon at bugs.documentfoundation.org]

https://bugs.documentfoundation.org/show_bug.cgi?id=135501

--- Comment #107 from John Mills <jmills59 at yahoo.com> ---
(In reply to Rafael Lima from comment #106)

Hi Rafael, thanks for taking the time to write such a reasoned response from
your perspective, this is entirely the type of feedback that we need, examining
the positives and negatives in a very important topic. I think your personal
experience is likely not too far from a lot of other people.

> Hi all! I have been using LibreOffice every day for a couple of years to do
> all my work and I would like to provide some insights on this issue.
> 
> At first glance, I would be totally in favor of switching to the Tabbed UI
> as the default option. On my computer I always switch to the Tabbed
> interface, because I really think it is more pleasing to my eyes. My
> preferred setup is LO using kf5 and dark mode with Tabbed UI.
> 
> With this setup, other people who are not LO users and see me using it with
> the Tabbed UI tend to compliment how good it looks. Since I often give
> lectures via online meeting services (Zoom, Google Meet, etc) and I share my
> screen, I had some people asking "What is this office suite you're using...
> it looks nice!". Most people don't even recognize it's LibreOffice, because
> they do not know we already support a ribbon-like interface.
> 
> I also agree with those who claim that, in order to attract new users we
> should provide them with something they're more familiar with, which is the
> ribbon interface. I have met people who migrated from LibreOffice to WPS /
> OnlyOffice / FreeOffice just because of the way they look and feel, as well
> as for the superior ribbon interface support (despite LibreOffice providing
> a much more feature rich application).

This is the critical part, the Tabbed UI provides an attractive (certainly on
Linux) and familiar interface to users coming from other office suites such as
MSO, OnlyOffice, Kingsoft, Softmaker to name a few. This is important, if you
were looking 50 to 10 years in to the future where do you think the desktop and
online office space is going to be? Will there be consolidation? Will the
desktop market shrink compared to online? Will MSO still be number one, will
new competitors enter the market? The fact is we don't know for certain, but if
trends continue then i think there will be an increased online presence and MSO
will still be the most popular desktop client. If they don't radically change
their UI then the 'ribbon' will be 20 + years old at that point and the type of
UI used by LO 30 years old. 

Just going by those numbers the current default UI paradigm used by LO will be
hopelessly out of date compared to the competition, we can see the movement
from traditional menu-based UIs now for a number of years in the office suites
mentioned.The very real concern is that this will cause users to migrate away
from from LO, free as in 'gratis' is the predominant concern for most users I
suspect and all of these other suites offer a product that is generally non
cost based to consumers, save for MSO however they offer mobile and web based
solutions without cost. I can sit a student down with LibreOffice using the
tabbed UI or Softmaker, Only Office and  WPS and have them being productive in
around 30 minutes once some of the differences are explained compared to using
MSO. If you do the same with the current UI then it is a lot longer and the
resistance factor is increased.   

The 'libre' in Libreoffice is critical of course, but is that the primary
factor that will increase or sustain the number of users over the next decade? 


> 
> HOWEVER, being a very frequent user of the Tabbed UI I have to recognize it
> has some serious limitations that make it unfit to be the default option.
> I'll list some of them:

Thank you for these criticisms, it is very important to hear the issues that
users have. And I agree with most of what you say here, but inless there is
emphasis and resources made available to correct these then nothing will happen
and that stagnation is not healthy for the LO application and community in the
longer term.
> 
> 1) The set of available commands in each Tab needs to be better thought out,
> since some commands are missing and some popular commands should be more
> prominent. This needs more UX research to improve the final layout. If we
> make the Tabbed UI the default (and it gets documented) we won't be able to
> keep changing it.

>From what I understand the structure of the tabs can be changed relatively
easily. There should be an emphasis for as little change as possible then over
the coming years. The standard lets say is MSO, but others like Kingsoft,
OnlyOffice and WPS use a similar layout so there is some learnings that can be
taken there. I would support in that matter if it was thought to be an exercise
to undertake. 

> 
> 2) The Tabbed UI does not respond well to window resizing because of the way
> the widgets are grouped. Sometimes, reducing the window by a small amount
> will hide half of the available commands.

Yes, indeed, I was showing Heiko this in the design meeting a few weeks back,
the reflow of icons in MSO is very good and logically thought out. Once
difference in MSO is that the icons next to the tabs (Save, Open, Undo, Redo,
Print) in LO are moved to the header bar in MSO this creates additional room
for the tabs themselves when resizing the window. I realise currently LO
doesn't support icons in the header bar but this would be an ideal way to get
back much needed horizontal space. 

> 
> 3) Some commands have disproportionally long names that use to much of the
> available width in the Tab, specially in translated versions of LO. This one
> is hard to fix, but it's a limitation of the Tabbed UI.

Yes, how is this handled in other Office Suites with simmilar UI paradigms, is
there some learnings that can be taken away?

> 
> 4) The Tabbed UI does not support much customization as the Standard UI does
> (see Tools - Customize - Notebookbar). Here the question is: do we want to
> support customization of the Tabbed UI? Maybe not, since it would introduce
> a whole lot of complexity.
> 

My broad opinion would be no, for a couple of reasons, mainly the documentation
side of things, but also there should be a vision of how this UI should work
across different OS versions and potentially the future for the web and
possibly mobile. I do think if header bar usage was possible in the future then
the ability to add icons there, similar to how MSO allows would be useful.

> 5) The Tabbed UI does not integrate well with extensions. The "Extension"
> tab is really hard to work with from the standpoint of extension developers
> and the merge commands do not work as expected. All I ever was able to do is
> add icons without labels. If the Tabbed UI is to become the standard, we
> need to improve the way extension developers will integrate their extensions
> into LO.

Absolutely, I understand that the extention bar came about through a Google
summer of code intern working on a project. There has been some debate for
example with the Zotero team on the applicability of this extension tab.
Currently it is not really fit for purpose and if the decision is that it is a
benefit then work is definitely needed there. 


> 
> 6) AFAIK the Tabbed UI does not integrate well with dark mode in Windows, so
> it would be a bad experience for Windows users.

I believe this is pretty much resolved now with the work that is being
undertaken with the dark mode in Windows, it would be worth you turning on this
experimental feature and provide feedback of your experiences. I have used it
for a number of months now and find the experience generally quite positive. 

> 
> 7) We need to agree on what we'll call this interface. Some people say
> "Notebookbar", "Notebook bar", "Tabbed UI"... IMO we need to settle with
> "Tabbed UI" to avoid confusion, because only experienced users will
> understand that Notebook bar and Tabbed UI is the same thing.

Tabbed UI is the simplest and most descriptive.


> 
> 8) All documentation (guides and help) assume the user is using the Standard
> UI, so we need to give time for the documentation team to provide the
> necessary changes.

Yes, I would expect that if any migration was done it would be over many months
at least allowing the documentation team to amend their documents.

> 
> 9) We should stick only with the Standard and Tabbed UI only, since the
> other UIs are not as well maintained. This would help us focus on the two
> most important options. Other variants could be provided as extensions for
> users.

Yes, agreed, more than two UIs can cause confusion, personally, I think rather
than removing them I would make them an option in the experimental section. I
do think there are some benefits to a compact version of the Tabbed UI however
and that could be accomplished potentially with an icon that collapses the tab
partly as is done currently in MSO.

> 
> In summary, I don't think we're near the point to make the Tabbed UI the
> default option. We need first to define a roadmap and improve it before
> considering a definitive switch.

A road map can only be defined if there is a decision that this is the correct
direction of travel for the project, clearly there is need for improvement as
you say. However if the resources are not made available then improvements
won't happen. There needs to be some strategic vision for where the UI needs to
be or else my concern is LibreOffice will simply stagnate and start to
hemorrhage users over the coming years.

-- 
You are receiving this mail because:
You are on the CC list for the bug.